Reactions of (dimethylamido) halophosphorus (III) with aluminum (III

Alan H. Cowley , Richard A. Jones , and Miguel A. Mardones. 1992,56-63 ... A. H. Cowley and R. A. Kemp. Chemical .... Steven A. Weissman , S. G. Baxte...
0 downloads 0 Views 757KB Size
3042 Inorganic Chemistry, Vol. 15, No. 12, 1976 (3) (a) See ref 2, Chapters 6-10; (b) ibid.,Chapter 9, pp 181, 190; also see p 143. (4) J. K. Johanneson, J . Chem. Soc., 2998 (1959). (5) W. A. Brungs, J . Water Pollut. Control Fed., 45, 2180 (1973). (6) J. D. Johnson and R. Overby, J. Sanit. Eng. Diu.,Am. Soc. Ciu. Eng., 97, 617 (1970). (7) H. Galal-Gorchev and J. C. Morris, Znorg. Chem., 4, 899 (1965). (8) J. D. Johnson and R. Overby, Anal. Chem., 41, 1744 (1969). (9) L. Farkas, M.Lewin, andR. J. Blwh,J. Am. Chem.Soc., 71,1988 (1949). (10) M. Lewin and M. Avrahami, J . Am. Chem. Soc., 77, 4491 (1955). (1 1) P. Engel, A. Oplatka, and B. Perlmutter-Hayman,J . Am. Chem. Sor., 76, 2010 (1954). (12) C. F. Prutton and S. H. Maron, J . Am. Chem. Soc., 57, 1652 (1935).

Kopp, Bond, and Parry H. A. Liebhafsky, and B. Makower, J . Phys. Chem., 37, 1037 (1933). (14) “Standard Methods for the Examination of Water and Wastewater”, 13th ed, American Public Health Association, Washington, D.C., 1971, pp 112-116. (1 5) I. M. Kolthoff and R. Belcher, “Volumetric Analysis”, Val. 3, Interscience, New York, N.Y., 1957, pp 449-473. (16) I. M. Kolthoff and J. J. Lingane, “Polarography”, Vol. 11, 2d ed, Interscience, New York, K.Y., 1952, p 946. (17) I . M. Kolthoff and K. H. Furman, “Potentiometric Titrations: A Theoretical and Practical Treatise”, Wiley, New York, N.Y., 1926. (18) I. M. Kolthoff and E. B. Sandeli, “Textbook of Quantitative Inorganic Analysis”, Macmillan, New York, N.Y., 1946, p 557. (19) C. M. Kelley and H. V. Tartar, J . Am. Chem. Soc., 78, 5752 (1956). (13)

Contribution from the Departments of Chemistry, University of Michigan, Ann Arbor, Michigan, 48 104, and University of Utah, Salt Lake City, Utah 84112

Reactions of (Dimethylamido) halophosphorus(II1) with Aluminum(II1) Chloride and with Some Etherates of Aluminum(II1) Chloride R. W. KOPP, A. C. BOND, and R. W. PARRY* Receiued March 22, 1976 AIC60226K The chloro(dimethylamido)phosphorus(III) ligands interact with AlC13 under appropriate conditions to give (CH3)2NPC12.AICI3, ((CH3)2N)2PCl.AlC13, and [2((CH3)2N)2PCI].AlC13. Tris(dimethylamido)phosphorus(III) and AIC13 react under comparable conditions to give ((CH&N)3P*AlCI3. The tris ligand will also react with ((CH3)2N)2PCI*AlCI3 to but it will not react with ((CH3)2N)jP-AlCI3 to give [2((CH3)2N)3P].AIC13. give ((CH3)2N)3P.((CH3)2N)2PCl.AICl3 Trimethylamine will gradually replace some of the phosphorus ligand from the complex to give some (CH3)3N*AlC13. A solution of dichloro(dimethylamido)phosphorus(III) ligand in isopropyl ether was stable as was a solution of aluminum chloride in isopropyl ether. On the other hand, an isopropyl ether solution containing A1C13 and the dichloro(dimethylamido)phosphorus(III) ligand gave 1 mol of isopropyl chloride for each mole of PC13 and each mole of the dichloro(dimethy1amido)phosphorus ligand. The ether solution turned red as a result of the formation of an aluminum chloride-olefin complex. The splitting process to generate alkyl halide was not observed with diethyl ether. The chloride transferred came from the phosphorus ligand, not the AlC13. The ligand ((CH3)2N)zPCI will displace about 70% of the diisopropyl ether from the diisopropyl etherate of AIC13 and ((CH3)2N)3P will displace 1OC% of the ether. Models to interpret these facts are presented.

The (dimethylamido)halophosphorus(III) ligands contain at least two quite different Lewis base centers. For example, the compound ((CH3)2N)2PF contains two potentially basic nitrogen atoms and one potentially basic phosphorus atom. The site utilized in a given acid-base process is dependent upon the nature of the acid used. When borane fragments such as BH3, B3H7, and R4Hg are used as the acid, bonding is known to be through the phosphorus at0m.I-j When other boron acids such as BF3, BCl3, or B(CH3)s are used, evidence indicates probable bonding through a nitrogen When aluminum alkyls are used as the acid, the behavior is more complicated. Clemens, Sisler, and Brey8 studied the reaction of (CH3)2NP(CH3)2 with Al(C2Hj)3. They reported that a P-A1 bond formed when the reagents were directly combined but that the structure changed to give an N-A1 bond when the adduct was heated. In this work the reactions of (dimethy1amido)halophosphorus(II1) ligands with AlC13 have been studied. Although it was originally assumed that a conventional acid-base adduct would be formed, structures and stoichiometry turned out to be quite different from those originally v i s ~ a l i z e d .In ~ this paper the appropriate chemistry is summarized. Subsequent reports will probe questions of structure and molecular dynamics. The System (CH3)2NPF2-AlC13 Initial work with (CH3)2NPFz and A1C13 revealed a halogen-exchange process similar to that reportedlo for the system F3P-AlCl3. To eliminate complications resulting from halogen *To whom correspondence should be addressed at the University of Utah.

interchange to give AlF3, the chloro ligand, (CH3)2NPC12, was used in place of its fluoro couterpart. The System (CH3)2NPC12-AlC13 When excess (CH3)2NPC12 and AlC13 were directly combined at 20 “ C in the vacuum system, the following reaction occurred almost quantitatively in 0.5 h or less 2(CH3),NPCl,

+ A1C13

--t

((CH,),N),PCI*AlCl,

+ PC1,

Even at a temperature of -23 “ C some PCl3 formed as soon as the reactants were brought into contact. The foregoing observations were quite unexpected, since the equilibrium constant for the disproportionation of the free ligand to give ((CH&N)2PCl and PCl3 has been estimated as 5 X and according to Van Wazer and Maier,’ an equimolar mixture of PCl3 and ((CH3)2N)zPCl undergoes essentially complete conversion to (CH3)2NPClz in less than 30 s at 25 OC; PCl3 is definitely not an expected product. An application of simple mass law arguments to the equation for the reaction of (CH3)2NPClz and A1C13 suggested the use of PCl3 as a solvent to minimize PCl3 formation and thus promote the formation of (CH3)2PClrAlC13. Indeed, the monoadduct (CH3)2NPC12*AlCl3 can be prepared from equimolar quantities of acid and base in PCl3 at -23 OC, over a 12-h period. The pertinent equation is

The solid 1:l adduct remaining after removal of the PCl3 solvent undergoes disproportionation upon standing at 25 “C. The products are ( ( C H ~ ) Z N ) ~ P C ~ - PC13, A ~ C and ~ ~ , A1C13.

Reactions of (Dimethylamido)halophosphorus(III)

Inorganic Chemistry, Vol. 15, No. 12, 1976 3043

phosphorus(II1) in isopropyl ether was stable under the conditions used, as was a solution of pure aluminum chloride in this ether. These facts prompted a more thorough study of the isopropyl etherate of aluminum chloride and of its (CH,), NPCl,*AlCl, + (CH,),NPCl, ((CH, ), N), PCl.AlC1, exchange reactions with (dimethylamido)halophosphorus( 111) t PCl, ligands. The Diisopropyl Etherate of Aluminum Chloride. The diWith trimethylamine a base displacement reaction occurs n-propyl etherate of AlCl3 was first described as a red liquid rapidly at 25 OC to give (CH3)2NPC12 and (CH3)3N.AlCl3. in 1915,13but the diisopropyl etherate of AlC13 prepared in The System ((CH3)2N)2PCbAICh this study was a white solid which showed no decomposition Direct combination of ((CH3)2N)2PCl and AlC13 in a 1:l over a 3-day period in the vacuum system at 25 OC. In the mole ratio gives ((CH3)2N)2PCLAlC13 as a waxy solid which presence of a trace of water it changed in a matter of a few is identical in properties with the material prepared from hours to a moist yellow solid. After a few days a red liquid (CH3)2NPClrAlCls and (CH3)2NPC12 as described above. covering a red-brown solid appeared. These observations can Reaction of ((CH3)2N)2PCl.AlC13 with trimethylamine gives be understood on the basis of literature records. A trace of ((CH3)2N)2PCl and (CH3)3N.A1C13 in a regular base diswater hydrolyzes the AlCl bond to liberate HC1. The HCl placement process. then attacks the ether to give some isopropyl chloride. The In a completely unexpected process ((CH3)2N)2PCbAlC13 resulting isopropyl chloride will split out HCl in the presence picks up a second mole of ((CH&N)2PCl to give a stable 2:l of AlCl3,l4 to give unsaturated hydrocarbons which complex adduct [2((CH3)2N)2PCl].AlC13. The compound ((CH3)2with the AlCb to give a colored solution. The scarlet complex N)2PCLAlC13 will also pick up 1 mol of ((CH3)2N)3P to give between AlC13 and an olefin was established by Burbage and ((CH3) 2N) 2PCl4 (CH3) 2N) 3P.AlC13. These complexes Garrett,ls who reported the evolution of HBr when liquid containing 2 mol of ligand/mol of AlCl3 were originally n-butyl bromide was added to AlBr3. A scarlet liquid addition formulated9 as five-coordinate aluminum species by analogy compound, 1-butene-aluminum bromide, was isolated and to the known12 structure of ((CH3)3N)yAlH3 and the characterized by Burbage and Garrett.ls Thus the develproposeds structure for ((C~H~)~P)~N(C~HS).A~(C;?HS)~ and opment of color in the higher etherates of aluminum chloride related species. Subsequent work has eliminated the fiveis a fairly sensitive criterion for ether decomposition. Further coordinate model in favor of an ionic one which is described discussion on the etherates of aluminum chloride is available in the accompanying paper. elsewhere. 16-18 Condensation of excess phosphorus(II1) ligand onto a 2:l Reaction of Phosphorus Trichloride with the Diisopropyl adduct such as [2((CH3)2N)2PCl]-AlC13simply dissolves the Etherate of Aluminum Chloride. Equimolar quantities of the complex. The N M R spectrum showed distinct peaks for both diisopropyl etherate of AlC13 and PCl3 were frozen at -196 the free and complexed ligand. No evidence for a 3:1 or higher OC and then warmed to 0 OC and held for 18 h in the vacuum adduct was found. line. The system turned red as expected. About a 1.07-mmol When Pc13 is condensed onto [2((CH3)2N)2PCl].AlC13, quantity of isopropyl chloride was separated for each millimole halogen-amido group exchange occurs as shown by of PCl3 used. When excess trimethylamine was added to the remaining product after removal of the R-C1, 1 mol of [2((CH3),N),PCl] *AlCl, + PC1, ((CH,),N),PCl*AlCl, (CH3)3N was retained for each mole of A1C13 originally t 2(CH,),NPCI, present, and (CH3)2CHP(O)C12 was released. The latter The compound of composition ((CH3)2N)3P.((CH3)2compound had an infrared spectrum which was identical with N)2PCLAlC13 fumes in moist air. It can be held at 65 OC for the spectrum of an authentic ~ a m p 1 e . l ~Equations which 3 days in the vacuum system with no evidence of decompodescribe the foregoing observations are sition. After 2 days at 98 OC in the vacuum system, P(NR2)3 ((CH,),CH),O~AlCl, t PCl, -+ (CH,),CHCl is driven off leaving a yellow residue. + (CH,),CHOPCl,~AICl, The System ((CH3)2N)3P-AIC13 (CH,),CHOPCL,~AlCl, + (CH,),N + (CH,),N*AlCI, An excess of ((CH3)2N)3P reacts with AlCl3 to give only + (CH,),CHOPCI, a one to one adduct, ((CH3)2N)3P*AlC13. No evidence for (CH,),CHCl a double adduct comparable to [2((CH3)fi2PCl]-AlCl3 was (CH,),CHOPCl, (CH,),CHP(O)CI, ever found. This fact is very difficult to harmonize with the Arbusov rearrangement original five-coordinate aluminum model for [2((CH3)2Very nearly 1 mol of chloride is transferred per mole of PCl3 N)2PCl].AlC13 and was one of the key observations leading used. The product (CH3)2CHOPC12 undergoes the expected to new structural postulates. Arbusov rearrangement in the presence of an alkyl halide. The Although ((CH3)2N)3P*AlC13will not pick up another mole red color of the product arises from the presence of the isoof ((CH3)2N)3P, it will pick up 1 mol of ((CH3)2N)2PCl to propyl chloride in the presence of AlC13 and the resulting give the previously identified compound ((CH3)2N)3P.((Cgeneration of a small amount of red olefin-aluminum chloride H3)2N)2PCl*AlC13. Addition of (CH3)zNPCh to ((CH3)Zcomplex. It is significant that no isopropyl ether was found N) 3P.AlCl3 appears to give halogen-amido interchange since in the products. the product is [2((CH3)2N)2PCl].AlC13. Similar halogenReaction of (Dimethylamido)dichlorophosphorus(III)with amido interchange is noted when PCl3 is added to ((Cthe Diisopropyl Etherate of Aluminum Chloride. When H3)2N)3P.AlCl3. The products recovered after 30 min are (CH3)2NPC12 was allowed to react with the diisopropyl equimolar quantities of (CH3)2N2PCbAlC13 and (CH3)zNetherate of aluminum chloride under conditions comparable PC12. to those used for Pcl3 (except reaction time was 7 h),19 nearly Reactions of the (Dimethylamido)halophosphorus(III) 1 mol of isopropyl chloride was formed for each mole of Ligands with Aluminum Chloride Etherates (CH3)2NPC12 used. Addition of excess (CH3)3N to the In early attempts to prepare amidohalophosphorus(II1) products gave 1 mol of (CH3)3N.AlCl3 for each mole of AlC13 adducts of AlCl3, diisopropyl ether was used as a solvent. originally used. As expected, the products were red, but no Significant quantities of PCl3 and isopropyl chloride were attempt was made to isolate the other expected substance, identified in the products. A solution of the (amido)halo((CH3)2CHO)P(NRz)Cl, or its rearrangement product. The

About 20-30% of the relatively pure adduct was found to decompose in 8 days at 25 "C. The solid also reacts rapidly at 25 OC with excess ligand +

+

+

3044 Inorganic Chemistry, Vol. 15, No. 12, 1976

Kopp, Bond, and Parry

equations describing the foregoing observations are

system at -196 "C. After warming of the mixture to 25 "C with stirring, the pure product (as determined by N M R : 6('H) -2.65 = 12.1 8(31P)-160 (H3P04)25)was obtained. (TMS),24JPNCH Reaction of Neat (CH3)2NPC12 with AICI3-Synthesis of ((CH3)2N)2PCI*AIC13and PC13. A typical reaction out of many is described. A 5.15-mmol sample of (CH&NPC12 was condensed onto 1.09 mmol of pure dry A1C13 in the vacuum system at -196 "C. The reaction vessel and contents were then warmed to 0 "C and held for 20 h with occasional stirring. Fractionation of the volatile components gave 2.91 mmol of unreacted (CH3)2NPC12 and 1.06 mmol of PC13. The waxy, water-sensitive solid remaining in the tube (mp 60-65 "C) had the overall formula ((CH3)2N)2PCI.AlC13. The detailed characterization of the compound is given under the description of the reaction of ((CH3)2N)2PCI and AIC13. In separate studies to determine rate of the reaction it was found that an equilibrium pressure of PCl3 was reached after 25 min at 25 "C. Reaction of (CH3)2NPC12 with AIC13 Dissolved in PC13-Synthesis of (CH3)2NPClpAIC13. In one run a 2.67-mmol sample of AlC13 was suspended in 6.530 g (47.51 mmol) of PC13. A 2.53-mmol sample of (CH3)2NPC12 was frozen at -196 "C above the AIC13-PC13 mixture. The system was warmed to -20 "C and held there for 9 h with occasional stirring. Fractionation of the volatiles pulled from the tube at -20 "C (-45 and -196 "C traps) produced 47.59 mmol of PC13 (-196 "C) and 0.28 mmol of (CH3)2NPC12 (-45 "C). When the system was warmed to room temperature, the product was a clear, fairly mobile liquid, perhaps from solvent PC13 generated by decomposition of the original product. As this liquid stood for 8 days, more PC13 was generated. An 0.44-mmol sample of PCl3 was pulled from the system. The product (CH3)2NPCl?AlCI3 is clearly unstable and converts readily to ((CH3)2N)2PCbAlC13 and PCl3. This instability proved to be somewhat of a problem in studying the dichloro(dimethylaniido)phosphorus(111)-aluminum chloride complex, but, as will be noted later, its decomposition could be followed and interpreted by use of N M R spectroscopy. When an excess of N(CH3)j (5.40 mmol) was condensed onto a 2.60-mmol sample of slowly decomposing (CH3)2NPC17AIC13 and the entire system was held for 18 h at 0 "C, volatile materials were obtained. Identified were 2.09 mmol of N(CH3)3 and 1.61 mmol of (CH3)2NPC12. The above facts suggest that about 60% of the original (dimethylamido)dichlorophosphorus(III) was displaced from the complex. The infrared spectra of the free ligand and the complex are summarized here. Assignments for the ligand are essentially those of Farran.26 The order used is as follows: frequency in cm-1, intensity (s = strong, w = weak, m = medium, v = very, sh = shoulder), and probable assignment (with v = stretch, p = rock, F = deformation). While the spectrum of the gaseous ligand is compared here to the spectrum of the solid complex, F a r r a n P comparison of the spectra of solid and gaseous (CH3)2NPC12 indicates that a comparison of gas and solid at this level of resolution is appropriate. As expected, the solid spectrum has additional lines. The spectrum for gaseous (CH3)2NPC12 is as follows: 2980 sh ( u c H , ) , 2935 m (vcH,), 2860 sh (YCH,), 2810 mw (vcH,), 1480 m ( ~ c H , ) , 1460 m ( ~ c H , ) , 1290 m (UCNP), 1187 m (PCH,), 1144 w (PCH,), 1068 mw (PCH,), 981 s (~cNP), 688 ms ( U N P ) , 530 ms (vpc~),508 vs (vpci), 456 s ( v p c ~ ) . For the solid complex (CH3)2NPCIpAlC13, the spectrum is as follows: 2960 sh (~cH,),2940 m (vcH,), 2800 w (VCH,), 1475 sh (FcH,), 1450 m (~cH,), 1410 sh (-), 1295 ms (UCNP), 1172 ms (PCH,), 1057 m (PCH,), 1009 sh ( U C N P ) , 984 vs (VCNP), 860 w (-), 805 w (-), 712 mw, ( v ~ p ) ,690 in (VNP), 641 m (-), 585 m, sh (upcl), 500 vs (AICI), 430 m, sh (-). Reaction of ((CH3)2N)2PCI with AICI3-Synthesis of ((CH3)2N)2PCI*AIC13. Equimolar amounts of resublimed AlC13 and ((CH3)2N)2PCI were combined in the vacuum line using previously described techniques. A waxy solid formed.27 The molecular weight of the product as measured by vapor pressure depression of CH2C12 at 0 "C was 265 f 35. The theoretical value for a monomer is 289. Additional characterization data are discussed under "structure" in a subsequent paper. The principal infrared frequencies of the free ligand were recorded by three separate investigators. Kopp recorded those of the neat liquid, Schultz recorded those of a CH2C12 solution, and Thomas recorded those of the liquid. In no case did the different spectra differ by more than ten wavenumbers. Lines were usually closer. The best values are summarized here using the same notation scheme that was outlined in the preceding section: 3000 w (vcH,), 2935 s (YCH,), 2900 s (vcH,), 2845 m sh (vcH,), 2805 m (vcH,), 1480 ms ( F c H J , 1450 s (~cH,), 1410 w sh (~cH,), 1280 s (UCNP), 1194 s (PCH,), 1160 sh (-), 1140 w (PCH,),

((CH,),CH),O~AICl, + (CH,),NPCl, (CH,),CHCl t (CH,),CHOP(N(CH,),)Cl*AlC13 (CH,),CHOP(N(CH,),)Cl~AlCl, t N(CH,), + (CH ,) ,N.AICl, t [ (CH ,) ,CHOP(N(CH ,)C1] +

Reaction of Bis(dimethylamido)chlorophosphorus(III) with the Diisopropyl Etherate of Aluminum Chloride. As expected from the foregoing result, the use of ((CH3)2N)2PCl with the diisopropyl etherate of aluminum chloride generated a very small amount of isopropyl chloride. About 70% of the original ether could be pumped slowly from the system, but evidence clearly suggested an unstable adduct involving the ether, the bis(dimethylamido)chlorophosphorus( 111),and the aluminum chloride. Reaction of Tris(dimethylamido)phosphorus(III) with the Diisopropyl Etherate of Aluminum Chloride. When ((CH&N)3P was mixed with ((CH3)2CH)20.AlC13 in a 1:l mole ratio at -196 OC, then warmed to 0 O C , and held at 0 OC for 5 h, complete displacement of the ether resulted. The other product was ((CH3)2N)3P*AlC13. As expected, no isopropyl chloride could be detected and no color was observed in the system. The foregoing process is a straightforwardbase displacement process and contrasts sharply with the chlorination process observed with (CH3)2NPC12 and with Pc13. In view of the foregoing facts it is concluded that the chloride of the alkyl chloride comes from the P-Cl, not the AlC13. Reaction of (Dimethylamido)chlorophosphorus(III) with the Diethyl Etherate of Aluminum Chloride. It is known that diisopropyl ether undergoes cleavage more easily than does diethyl ether;21 thus the reaction of (C2H5)20.AlC13 and (CH3)2NPC12 was of interest. When an equimolar, homogeneous mixture of (C2H5)20#AlC13and (CH3)2NPC12 was held at 0 "C for about 10 h and then fractionated, about 11% of the original ether was displaced and about 89% of the original phosphine was recovered. Only a trace of ethyl chloride was detected. As expected, the halide-transfer process is less facile with the diethyl etherate than with the diisopropyl etherate. Experimental Section General Procedures. Standard vacuum-line proceduresz2 were used throughout. Storage tubes and reaction tubes were equipped with greaseless Fischer-Porter 4-mm Teflon needle valves and were connected to the vacuum system with O-ring joints. Apiezon grease was used on other glass valves. All solid materials were handled under dry N2 in a drybox. Molecular weight estimates were made by dissolving a known amount of the sample (about 300 mg) in a known amount of CH2Clz (about 8 8). Vapor pressure at 0 "C was measured. Infrared spectra were taken on a Perkin-Elmer 337 grating spectrometer. For gases a cell with KBr windows was used. Liquids were run as films on a KBr plate; solids were run in a KBr pellet. Nuclear magnetic resonance data were obtained using a Varian A-60, or a Varian HR-100 spectrometer. Detailed spectra, which provided the basis for all structural assignments, will be presented in the following paper. Starting Materials. Commercial AlCI3 (Baker and Adamson reagent grade) was purified by multiple sublimation in a stream of dry nitrogen. Malinckrodt reagent grade PC13 was condensed onto dry AlCl3 and then fractionated through traps at -45, -83, and -196 "C. The -83 " C fraction was used. The (CH3)3N was generated from (CH3)3NHC1 (Eastman) using 50% NaOH. The material was passed through a -83 "C trap into a -196 "C trap. The -196 "C fraction was used. Commercial reagent grade ethers were dried over CaH2. Methylene chloride was stored over Drierite. The (CH&NPC12 was prepared by adapting the procedure of Burg and S l ~ t toa vacuum-line ~ ~ techniques. Details are available e l ~ e w h e r e . ~ The ((CH3)2N)3P was prepared by an adaptation9 of the method of Holmes and Wagner.6 The ((CH3)2N)2PC1 was prepared by combining 2 mol of ((CH3)2N)3P with 1 mol of Pc13 in the vacuum

Reactions of (Dimethylamido)halophosphorus(III)

Inorganic Chemistry, Vol. 15, No, 12, 1976 3045

Table I. Reaction of the Diisopropyl Etherate of AlCl, with 1095 vw (-), 1062 m (PCH~),980 sh ( V C N P ) , 960 vs ( V C N P ) , 880 mw ((CH3)zN)nPC13-n (n = 0-3) (-), 790 w (-1,700 ms (VPN), 672 m (VPN), 510 sh (YPCI),490 m (VPCI), 410 w (~cNc), 307 w (-). Amt, mmol For the solid complex (( C H ~ ) Z N ) ~ P C ~ . A the I C Ifrequencies ~ and LiEther- Time: RC1: Ether: . w assignments are as follows: 2930 s (vcH,), 2815 m ( v c H ~ )2420 Lieandused nand ate h comulex complex (-), 1585 w (-), 1475 sh and 1450 s ( ~ c H , ) , 1410 w (~cH,), 1309 s (VCNP), 1220 sh (VCNP), 1186 s (PCH~). 1135 sh (-), 1003 sh (VCNP), 18 1.07 Trace 3.14 3.10 PC1, 996 vs (VCNP). 805 w (combn), 733 w (combn), 710 vw (VPN), 620 7 1.0 Small 7.81 7.87 (CH3)z NPClz W (VPN), 495 VS ( V A k d , 437 W, sh (6CNC). amt” Reaction of ((Ck3)2N)2PCI with ((CH3)2N)2PCl.AICI3-Synthesis ((CH,),N),PCl 3.63 3.53 18 Small 0.69 of the Diadduct [Z((CH3)2N)2PCl].AIC13. If ad excess of ((Camt HqbNbPCI (3.83 mmol) is condensed onto pure AICl3 (0.78 mmol) ((CH,)z N)3 P 1.53 1.52 5 0 1.0 - - ,. and allowed to react at 0 OC for 18 h as previolisly described, the ratio a In all cases reagents were mixed at -196 “C and warmed to of ligand to AlCI3 can be reduced to 2.15: 1 by exposing the reaction 0 “C, where temperature was held for the time indicated. vessel to the vacuum pumps. By flooding the product with hexane Treatment of the residue with (CH,),N resulted in pickup of and pumping it off several times, we can reduce the ratio to 2.06:l. 1 mol of (CH,),N/mol of AlC1, and release of ((CH,),CH)P(O)Cl, A pure product can be obtained by carefully combining the reagents as determined by infrared analysis. in a 2:l ratio. The solid white product melts over the range 118-123 “C; it is The resulting dry white solid was stable for at least 3 days at room soluble in methylene chloride and reacts vigorously with water vapor temperature. No evidence of isopropyl chloride or red color was found and almost explosively with liquid water. It has low solubility in if reagents and system were scrupulously clean and dry. benzene, hexane, chloroform, bromoform, and carbon tetrachloride. Reaction of ((CH3)2cH)20AICI3with ((CH3)2N)nPa3-n n = 0-3). It molecular weight as determined by vapor pressure depression of All reactions were carried out as described above. Results are CHzClz at 0 O C was 430 & 60. A monomer would have the theoretical summarized in Table I. value of 443. The infrared spectrum of the solid 2:l adduct Reaction of (CH&NPF2 with AlCI3. When (CH3)zNPFz and [2((CH3)2N)2PCI].AlCl3 is given here using the same notation as A l a 3 were mixed with a slight excess of ligand, a complex process that used in the preceding section: 2935 ms (vcH,), 2900 m (vcH~), occurred. About two-thirds of the (CH3)2NPF2 appeared as 2850 w (vcH,), 2810 mw (VCHJ, 1475 m sh ( ~ c H , ) , 1455 m ( ~ c H ~ ) , (CH&NPC12 and one-sixth as PF3. In a typical run a 1.59-mmol 1410 w sh (BcHJ, 1300 s (VCNP), 1170 s (PCH~), 1140 w (PCH,), 1062 sample of A1C13 and a 0.37-mmol sample of (CH&NPF2 were S (PCH,), 1008 S sh (UCNP), 984 VS (VCNP), 745 mw (-), 692 W (VNP), condensed in the vacuum line at -23 OC. After 15 min the reaction 668 m (VNP), 490 vs (VAICI), 437 w ( ~ c N c ) . mixture could not be condensed at -80 O C indicating the presence Reactions of ((CH3) 2N)2PCI*AICI3 with Various Reagentsof PF3. ‘A total of 3.69 mmol of (CH3)2NPF2 was added in three Synthesis of ((CH3)2N)2PCI*((CH3)2N)3P*AICl3. In general, reagents increments at -23 “ C over a period of about 1 h; then the mixture were condensed onto a sample of ((CH3)2N)2PClAC13 in the vacuum stood for about 4 h at -23 OC. Fractionation of the products indicated system at -196 OC. The temperature was raised to 0 OC and held that for 1.59 mmol of 2.175 mmol of (CH3)zNPFz was for periods ranging from hours up to 5 days. Volatile materials were consumed. A 1.369-mmol sample of (CH3)2NPC12 and a 0.24-mmol then removed at 0 OC and fractionated. sample of PF3 were obtained. An excess of trimethylamine displaced less than half of the ligand Acknowledgment. Support of this work by the National in a single exposure. Removal of volatile materials followed by Science Foundation through Grant GP 32079 to the University reexposure to the amine permitted recovery of up to 68% of the original ((CH3)2N)zPCl in the complex. of Utah and by a NASA Traineeship and Proctor and Gamble A fourfold molar excess of ((CH&N)3P was allowed to stand 5 Fellowship to RWK at the University of Michigan is gratefully days on the complex at 25 “C. When excess ((CH3)2N)3P was acknowledged. pumped off, 1.07 mol of ((CH&N)3P was retained for each mole Registry No. ((CH3)2N)2PCbAIC13, 60607-14-9; (CH3)zNPof complex used. C12.AIC13, 60594-92-5; [2((CH3)2N)2PCl].AlC13, 60594-85-6; Reaction of ((CN3)2N)3P with AI&. In a typical run carried out ((CH3)2N)2PCl-((CH3)2N)3P-AlC13,60594-83-4; ((CH3)2N)3P*AlC13, as described earlier, a 2.11-mmol sample of reacted with 2.19 60594-91-4; ((CH3)2CH)20,AIC13, 18785-84-7; (CH3)2NPC12, mmol of ((CH&N)3P. The complex appeared as an insoluble 683-85-2; AIC13,7446-70-0; ((CH3)2N)2PCl, 3348-44-5; (CH3)2NPF2, precipitate in liquid ligand if excess ligand were used. The adduct 8 14-97-1; PCl3, 77 19- 12-2. melted at 105-1 12 OC and then decomposed to a dark brown solid. The molecular weight of the solid in methylene chloride (mole fraction References and Notes of solute 0.127) at 0 “C was 320 & 50. The theoretical value for a (1) (a) S. Fleming, Ph.D. Dissertation, University of Michigan, Ann Arbor, monomer is 297.5. It was of special significance that no evidence Mich., 1963; (b) S. Fleming and R. W. Parry, Inorg. Chem., 11, 1 (1972). for a complex with 2 mol of ligand/mol of A m 3 was ever found. (2) M. D , LaPrade and C. E. Nordman, Inorg. Chem., 8, 1669 (1969). Reaction of the complex with ((CH&N)zPCl gave the previously (3) H. Noth and H. J. Vetter, Chem. Ber., 96, 1298 (1963). (4) R. G. Cavell, J . Chem. SOC.,1992 (1964). described double adduct: ( (CH~)~N)~P.((CHJ)~N)~PC~-AIC~~. (5) T. Reetz and B. Katlafsky, J . A m . Chem. SOC.,82, 5036 (1960). Condensation of a 3.00-mmol sample of (CH3)2NPC12 onto 1.22 (6) R. H. Holmes and R. R. Wagner, J . Am. Chem. SOC.,84,357 (1962). mmol of ((CH3)2N)3P*AlC13at -196 O C followed by warming to room (7) The existence of a B-N bond is less certain than the existence of the temperature gave evidence for the absorption of 1.12 mmol of B-P bond, but data are still fairly convincing. (CH3)2NPC12. The facts are best summarized by the equation (8) D. F. Clemens, H. H. Sisler, and W. S. Brey, Inorg. Chem., 5,526 (1966). (9) R. W. Kopp, Ph.D. Dissertation, University of Michigan, Ann Arbor, ((CH,),N),P~AlCl, (CH,),NPCl, --* [2((CH,),N),PCl] -AlCl, Mich., 1966. (10) E. R. Alton, R. G. Montemayor, and R. W. Parry, Inorg. Chem., 13, When a large excess of PCl3 (32.98 mmol) was condensed as before 2267 (1974). onto a 2.1 1-mmol sample of ((CH3)2N)3P*AlC13, two liquid layers (11) J. R. Van Wazer and L. Maier, J . Am. Chem. SOC.,86, 811 (1964). were visible about 30 min after the system had been warmed to room (12) C. W. Heitsch, C. E. Nordman, and R. W. Parry, Inorg. Chem., 2,508 temperature. After 3 days fractionation of volatile products gave 1.76 (1963). (13) G. B. Frankforter and E. A. Daniels, J. Am. Chem. SOC.,37,2560 (1915). mmol of (CH3)2NPC12 and 30.98 mmol of PCl3. The best equation (14) E. Wertyporoch and T. Firla, Z . Phys. Chem., Abt. A , 162, 398 (1932). to summarize these facts is (15) J. J. Burbage and A. B. Garrett, J . Phys. Chem., 56, 730 (1952). ((CH3),N),P*A1Cls + PCI, --* ((CH,),N),PCl*AlCl, (16) F. Johnson, “Friedel-Craftsand Related Reactions”, Vol. 4, G. Olah, Ed., Interscience, New Yoork, N.Y., 1963. (CH,),NPCI, (17) G. Urry, “The Chemistry of Boron and Its Compounds”,E. Muetterties, Ed., Wiley, New York, N.Y., 1967, p 332. Reaction of AICl3 with Isopropyl Ether. A 37.96-mmol sample of (18) P. Hagenmuller and J. Rauxel, C. R. Hebd. Seances Acad. Sci., 247 isopropyl ether was condensed a t -196 O C onto a 1.45-mmol sample (1958). of A1C13. The system was warmed to 25 “ C and the resulting (19) A. M. Kinnear and E. A. Perren, J . Chem. SOC.,3437 (1952); L. J. suspension was stirred for 6 h. When excess ether was removed, a Bellamy, “The Infra-red Spectra of Complex Molecules”, Wiley, New York, N.Y., 1957, p 314. 1.45-mmol sample of AlC13 retained a 1.45-mmol quantity of ether.

+

+

C. W:Schultz and R. W. Parry

3046 Inorganic Chemistry, Vol. 15, No. 12, 1976 (20) Under the conditions used here about 11% of the original (CH3)2NPC12 was recovered unreacted. Depending on time and temperature this value varied from 0 to 25% of the amount used. (21) As pointed out in ref 16 an increase in the electron-releasing power of the R group of the ether facilitates ether cleavage. The isopropyl group is usually considered to be more electron releasing than the ethyl group. (22) D. F. Shriver, "The Manipulation of Air-Sensitive Compounds", McGraw-Hill, New York, N.Y., 1969. (23) A. B. Burg and J. Slota, J . Am. Chem. Soc., 80, 1107 (1958). (24) H. J. Vetter, Naturwissenschuften, 51, 240 (1964). (25) J. K. Ruff, J . Am. Chem. Sac., 83, 2835 (1961). (26) C. F. Farran, Ph.D. Dissertation, University of Michigan, Ann Arbor, Mich., 1966. Farran's high-resolution spectrum of solid (CH3)2NPC12 and his assignments are as follows: 3000.8 s (v"cH,), 2993.4 s (v"cH,).

2942.4 s ( ~ ' c H , ) . 2922 m ( u ' c H ~ ) .2900.4 s (VICH,), 2878.3 m and 2864.9 m, sh ( v ' c H ~ )2841.2 , s (-), 2809.4 s and 2796.7 m (u'cH,), 1991.2 w (1303 696), 1680.6 w (1303 392), 1487.2 s ( ~ " c H ~ )1463.1 , sand 1451.4 s (~"cH,), 1443.2 sh (-1, 1440.8 s (GcH,), 1432.2 m (~'cHJ, 1429.3 m (-), 1402.0 s (~'cH,), 1303.2 s a n d 1298.3 sh (v'cN), 1188.3 m (696 + 498), 1174.4 s (~'cH,), 1138.6 w ( ~ " c H J , 1096.9 ms (-), 1065.7 sand 1063.9 s (~"CH?). 1029.5 w (696 335), 984.4 vs (u'cN), 833.7 vw (498 + 335), 731 vw (393 + 335), 696.5 s ( u ' P N ) , 498.0 vs ( u ' p c ~ ) ,437.2 vs (~"pci), 392.5 vs ( ~ " N C ~ )335.2 , s ( ~ ' N c ~292.6 ), w (-), 220.2 m (Y'PN). (27) In an improved process developed by M. Thomas (Ph.D. Dissertation, University of Utah, Salt Lake City, Utah, 1975), it was found that a 15% excess of AlCl3 eliminated formation of any contaminating 2 1 adduct. The product can be purified by crystallization from CH2C12 at -10 O C . (28) Not isolated or identified; may rearrange.

+

+

+

Contribution from the Departments of Chemistry, University of Michigan, Ann Arbor, Michigan 48104, and University of Utah, Salt Lake City, Utah 84112

Structure of [2( ( C H 3 ) 2 N ) 2 P C l ] -AlCl3, ( ( C H 3 ) 2 N ) 3 P - ( (CH3)2N)2PCl-AlC13, and Related Species-Diphosphorus Cations C. W. SCHULTZ and R. W. PARRY* Received March 26, 1976

AIC60237D The bis-ligand adducts of AICl3, [2((CH3)2N)zPCl].AlC13and ((CH3)2N)3P.((CH3)2N)2PCbAlCl3, form easily but have been difficult to describe in conventional acid-base formalism since the analogous bis adduct containing 2 mol of ((CH&N)3P/mol of AIC13 could not be prepared. Structures of all pertinent compounds now provide an explanation. The compound [2((CH3)2N)2PCl].AlC13is shown to have the structure [((CH3)2N)3P-P(N(CH3)2)Cl]+[AlCI4]-, while ((CH~)~N)~P.((CH~)~N)ZPCI.A~CI~ has the structure [((CH3)2N)3P-P(N(CH3)2)2]+[AlCl&. These conclusions are based on conductivity data, NMR spectra for 'Hand 31Pnuclei, infrared spectral data, and chemical information. The compound ((CH3)2N)3P*AlCI3which will not add additional ((CH3)2N)3P is shown to be a dimer with the initial structure [((CH3)2N)3P-P(N(CH3)2)2]+[NR2(AlCl3)2]-. The anion does not appear stable but undergoes further rearrangement. Apparent differences in chemistry are really due to the fact that the anion [R2NAIC13]+picks up a second mole of AlC13 very easily. The [GaC14]- and [PF& salts of the diphosphorus cation have been prepared, but we were not able to prepare [((CH3)2N)3P-P(N(CH3)2)F]+ in which a fluoride ion is coordinated to one phosphorus atom of the cation. In the first paper of this series' interactions between (dimethylamido)chlorophosphorus(III) ligands and aluminum chloride were described. Conventional one to one addition compounds of the general formula ((CH3)2N),C13-nP.AlC13 were reported along with two unexpected two to one addition compounds of formulas [2((CH3)2N)2PCl].AlC13 and ((CH3)2N)3P-((CH3)2N)2PCl.AlCl3.Since both of these compounds were quite stable if air and water were excluded, it was particularly puzzling that no two to one addition compound involving the closely related ((CH3)2N)3P and AlC13 could be prepared. All attempts to prepare [ 2 ((CH3)3N)3P] .AlC13] have been singularly unsuccessful. Structural answers to correlate the foregoing facts are provided in this paper. While the original structural questions seemed to center around the matter of a phosphorus-aluminum bond vs. a nitrogen-aluminum bond, it soon became clear that models based on either premise were unsatisfactory. In contrast, an ionic representation analogous to the cation [ (CH3)3P-P(CH3)2]+ reported by Noth2a and related species reported by Summers and Sisler2bgave a satisfactory representation of all observed properties. The compound [2((CH3)2N)2PCI].AlC13] is best described by structural formula I. The related

J I *Author to whom correspondence should be addressed at the University of Utah.

compound ((CH3)2N)3P-[(CH3)2N)2PCl-AlCl3 is best described by structural formula 11. The evidence to support

[

y(CH3)2

y(CH3)z

(1) (i)

(CH3),NP

P:

~

N(CH312

N(CH,L

I1

these assertions follows. Conductivity Data If an ionic formulation is a proper representation of the foregoing compounds, solutions of each substance in methylene chloride should be electrical conductors. This expectation has been verified. A M solution of AlC13 in methylene chloride has an equivalent conductance of 0.21 f 0.01 Si-l cm2 equiv-I. A comparable M solution of ((CH3)2N)2PC1 has an equivalent conductance of less than 0.001 C1cm2 equiv-l. On the other hand, a 0.3 M solution of [2((CH3)2N)2PC1].AlC13 has an equivalent conductance of 22.2 f 0.3 Si-l cm2 equiv-I. A simple ionic halide reference such as a 0.3 M solution of [(CsH5)4P]Cl in methylene chloride gave a value of 2.7 f 0.3 Si-* cm2 equiv-', and a reference covalent system such as C6H4Br2 gave a conductance of 0.0008 f2-l cm2 equiv-1 under comparable conditions. Clearly an ionic representation is appropriate.3 In the first paper of this series' a molecular weight value of 430 f 60 was obtained from vapor pressure depression measurements in CH2Cl2, while a theoretical value of 443 was expected for [2((CH3)2N)2PCl].AlC13. The data would be consistent with a mildly dissociated ionic system (Structure I) in methylene chloride.